Jump to content


Photo

SimulationCraft, a tool Mograin, Rivendare, Gorefiend and Thassarian would have used


  • This topic is locked This topic is locked
110 replies to this topic

#41 Starfox

Starfox

    King Hippo

  • Members
  • 578 posts

Posted 02 December 2009 - 05:26 PM

Is it possible for you to run the CLI? The GUI is running with Qt+MinGW and the CLI is using MSVC++.

Running on my development platform (Linux) using the profile above I get a nice tight result with minimal variance. I am unable to test it on Windows until much later tonight.

To use the CLI, take your profile and add the following to the top:
patch=3.3.0
calculate_scale_factors=1
iterations=10000

If you want to speed it up, you can just analyze a subset of the stats:
scale_only=str,agi

simc.exe your_profile output=output.txt html=output.html



SimulationCraft for World of Warcraft release 3.2.2 ( iterations=999, max_time=300, optimal_raid=1, smooth_rng=0 )
http://chart.apis.go...000,20&noop.jpg
>:C
Hello.
Light the fuse.
For all my homies.
Do not run, we are your friends.
SimulationCraft Druid Guy

#42 tripp6sic6

tripp6sic6

    Von Kaiser

  • Members
  • 31 posts

Posted 02 December 2009 - 07:40 PM

There is much less variance between runs with the CLI, but there are still some stats that have major outliers. Is this just a statistical problem and require more iterations? Here is the data that I have.

Using the CLI in Windows with the same profile, but with the default 20k iterations.

1.9901 1.1468 0.7071 1.7490 0.6151 2.4641 1.7375 1.2576 4.8160
2.0010 1.1403 0.7076 1.7060 1.3537 2.4700 1.7374 1.2739 3.6265
1.6122 1.1358 0.7022 1.7828 1.3260 2.4531 1.7275 1.2589 3.6601

Using the CLI in Windows with the same profile, but with 10k iterations.

1.9965 1.1733 0.7177 1.7291 1.3698 2.4444 1.7197 1.2701 4.7621
1.9802 1.1676 0.7079 1.7825 1.3504 2.4781 1.3668 1.2622 4.7899
1.9946 1.1685 0.7012 1.7727 1.3584 2.4738 1.7532 1.2432 4.7134

#43 dedmonwakeen

dedmonwakeen

    Bald Bull

  • Members
  • 1,302 posts

Posted 02 December 2009 - 07:55 PM

There is much less variance between runs with the CLI, but there are still some stats that have major outliers. Is this just a statistical problem and require more iterations? Here is the data that I have.

Using the CLI in Windows with the same profile, but with the default 20k iterations.

1.9901 1.1468 0.7071 1.7490 0.6151 2.4641 1.7375 1.2576 4.8160
2.0010 1.1403 0.7076 1.7060 1.3537 2.4700 1.7374 1.2739 3.6265
1.6122 1.1358 0.7022 1.7828 1.3260 2.4531 1.7275 1.2589 3.6601

Using the CLI in Windows with the same profile, but with 10k iterations.

1.9965 1.1733 0.7177 1.7291 1.3698 2.4444 1.7197 1.2701 4.7621
1.9802 1.1676 0.7079 1.7825 1.3504 2.4781 1.3668 1.2622 4.7899
1.9946 1.1685 0.7012 1.7727 1.3584 2.4738 1.7532 1.2432 4.7134


Since I am seeing variance "spikes" in your data as opposed to just "spreads" I am inclined to believe there is a bug rattling around in there. I'll take a peek over Starfox's shoulder......

#44 Starfox

Starfox

    King Hippo

  • Members
  • 578 posts

Posted 02 December 2009 - 09:39 PM

Seems like we found the cause, uninitialized variable :o
Hello.
Light the fuse.
For all my homies.
Do not run, we are your friends.
SimulationCraft Druid Guy

#45 tripp6sic6

tripp6sic6

    Von Kaiser

  • Members
  • 31 posts

Posted 02 December 2009 - 10:33 PM

I have found a bug with my 3.3 Unholy profile (above). It seems that Blood Strike and Plague Strike are being treated as if they are attacking from the front, but only those abilities. There is a roughly 7.7% parry rate on those abilities.

#46 Starfox

Starfox

    King Hippo

  • Members
  • 578 posts

Posted 02 December 2009 - 10:52 PM

I have found a bug with my 3.3 Unholy profile (above). It seems that Blood Strike and Plague Strike are being treated as if they are attacking from the front, but only those abilities. There is a roughly 7.7% parry rate on those abilities.


Yea, just noticed that too, was a major fuckup on my side:
ToT: the OH hit cannot have miss or dodge as a result, so I disabled MISS/DODGE before the OH hit and re-enabled it afterwards
And somehow I managed to slip PARRY in there too >:<
Fixed in latest revision, also made a run with all 4 DK profiles and updated the first post (DRW still not working, do I really have to c/p the whole dk module into the pet itself and add *0.5? X: )
Hello.
Light the fuse.
For all my homies.
Do not run, we are your friends.
SimulationCraft Druid Guy

#47 Darkside

Darkside

    I find your lack of faith disturbing.

  • Allied Members
  • 9,820 posts

Posted 02 December 2009 - 11:02 PM

Fixed in latest revision, also made a run with all 4 DK profiles and updated the first post (DRW still not working, do I really have to c/p the whole dk module into the pet itself and add *0.5? X: )


Couldn't you just add a flag to all the applicable attacks (HS, PS, IT, DC, DS, FF and BP) to apply a second time w/o talent bonuses (and at 50% strength) when the DRW is active?

ginger booty get on with yo bad self


#48 dedmonwakeen

dedmonwakeen

    Bald Bull

  • Members
  • 1,302 posts

Posted 03 December 2009 - 03:20 AM

Couldn't you just add a flag to all the applicable attacks (HS, PS, IT, DC, DS, FF and BP) to apply a second time w/o talent bonuses (and at 50% strength) when the DRW is active?


The DoTs complicate matters a bit. To be honest it isn't that hard to create a Mini-Me sub-class of death_knight_t.

Mini-Me can create an instance of each type of action. When the Master performs an action, he triggers the corresponding Mini-Me action. Since Mini-Me doesn't have any talents specified, you get that reduction for free. I've been meaning to add the player-level damage multiplier anyway, so we could integrate the 50% modifier there.

The harder part is associating the damage with the Master. The list of pets assumes a list of pet_t pointers. I'll see what breaks if I change that to player_t.

#49 dedmonwakeen

dedmonwakeen

    Bald Bull

  • Members
  • 1,302 posts

Posted 03 December 2009 - 06:23 AM

Seems like we found the cause, uninitialized variable :o


SimC v14 just released. It includes (among other things) the fix for this nasty DK bug.

#50 Darkside

Darkside

    I find your lack of faith disturbing.

  • Allied Members
  • 9,820 posts

Posted 03 December 2009 - 08:28 AM

In this new code, Blood Plague no longer crits under the 3.3 rules, while Frost Fever still does. This is the opposite of the behavior on ptr. Stat weights have normalized themselves though!

I don't want to be seen as a guy beating a dead horse, but the numbers for Obliterate are still way off.


e: I'm noticing a dramatic difference between weights generated with smooth-rng and those without.

Example:

w/o smoothing
2.6418 1.4269 1.0000 1.9301 1.7699 0.8355 1.6539 1.7012 6.4798

with smoothing
2.9873 1.2276 1.0000 3.1468 1.6395 1.9617 1.7583 1.6353 7.0370

ginger booty get on with yo bad self


#51 Starfox

Starfox

    King Hippo

  • Members
  • 578 posts

Posted 03 December 2009 - 10:47 AM

In this new code, Blood Plague no longer crits under the 3.3 rules, while Frost Fever still does. This is the opposite of the behavior on ptr. Stat weights have normalized themselves though!

I don't want to be seen as a guy beating a dead horse, but the numbers for Obliterate are still way off.


e: I'm noticing a dramatic difference between weights generated with smooth-rng and those without.

Example:

w/o smoothing
2.6418 1.4269 1.0000 1.9301 1.7699 0.8355 1.6539 1.7012 6.4798

with smoothing
2.9873 1.2276 1.0000 3.1468 1.6395 1.9617 1.7583 1.6353 7.0370

Oops, always thought it was FF that stayed on the setbonus. Fixed in r4047

Obliterate
80% normlized wpndmg +467 (+336 awareness)
*= 1 + 0.2 *glyph
*= 1 + 0.04 *merciless_combat
*= 1 + 0.125*disease_count
crit% += subversion * 3% + rime * 5%
crit_bonus_multiplier *= 1.0 + 0.15 * GoG

That's what it looks like now
Hello.
Light the fuse.
For all my homies.
Do not run, we are your friends.
SimulationCraft Druid Guy

#52 Guest_Afabar_*

Guest_Afabar_*
  • Guests

Posted 03 December 2009 - 12:09 PM

Merciless Combat is 6% per talent point.

#53 Starfox

Starfox

    King Hippo

  • Members
  • 578 posts

Posted 03 December 2009 - 12:55 PM

Merciless Combat is 6% per talent point.


Whoops, fixed

Still doing testing with DRW, its HS it is not 50% of your damage.
0/5 Blood Gorged, 0%,
3/3 Bloody Strikes

3 Bloody Vengeance stacks
UP
Undead (no 5% orc racial)
Dummy amor reduces damage: ~0.5887

4-6 dmg 2h, 2601 AP
DK HS:
~655 = 0.5887 *(5+ 2601/14*3.3+736) *.5 *1.45 *1.09 *1.04
Meets perfect, most hits are 655, with a 656 here and there

Now you would think that DRW would just hit for:
~327 = 0.5887 *(5+ 2601/14*3.3+736) *.5 *1.45 *1.09 *1.04 *0.5
But that is way to low
~475 = 0.5887 *(5+ 2601/14*3.3+736) *.5 *1.45 *1.09 *1.04 *0.5*1.45
It just seems as Bloody Strike get applied twice for DRW

Ok, so it seems that DRW has its own weapon, it does NOT copy the dk weapon.
Hello.
Light the fuse.
For all my homies.
Do not run, we are your friends.
SimulationCraft Druid Guy

#54 tripp6sic6

tripp6sic6

    Von Kaiser

  • Members
  • 31 posts

Posted 03 December 2009 - 06:06 PM

Is the increase in Bone Shield duration for 3.3 active? I don't see it in the source.

Is there a good way to make Bone Shield cast either pre-combat or in combination with Blood Tap with a conditional for how much time is left on the buff?

Also, I found a bug that I don't think is DK specific (I will test to make sure), but it seems that profiles that use the Heroic and non-Heroic versions of Death's Choice/Verdict will only use one proc. The workaround I have been using is to use the alternative name for one of the trinkets.

#55 Starfox

Starfox

    King Hippo

  • Members
  • 578 posts

Posted 03 December 2009 - 06:27 PM

Is the increase in Bone Shield duration for 3.3 active? I don't see it in the source.

Is there a good way to make Bone Shield cast either pre-combat or in combination with Blood Tap with a conditional for how much time is left on the buff?

Also, I found a bug that I don't think is DK specific (I will test to make sure), but it seems that profiles that use the Heroic and non-Heroic versions of Death's Choice/Verdict will only use one proc. The workaround I have been using is to use the alternative name for one of the trinkets.

Death's choice determines the proc by the the AP the trinket has: 288 vs 256, both proc independantly
Bone Shield 5mins is in with r4050
Also, if you cast Bone Shield as one of your first actions while you aren't in combat, it won't produce RP or consume runes.
But the duration and cd is lowered by 8-16s (random amount in that range) so you can simulate casting it a few seconds before pull

Synching actions can be done, I think it would be:
actions+=/blood_tap,sync=bone_shield,if=in_combat
actions+=/bone_shield,if=!in_combat|buff.bone_shield.remains<=X

Hello.
Light the fuse.
For all my homies.
Do not run, we are your friends.
SimulationCraft Druid Guy

#56 Darkside

Darkside

    I find your lack of faith disturbing.

  • Allied Members
  • 9,820 posts

Posted 03 December 2009 - 08:03 PM

Oops, always thought it was FF that stayed on the setbonus. Fixed in r4047

Obliterate
80% normlized wpndmg +467 (+336 awareness)
*= 1 + 0.2 *glyph
*= 1 + 0.04 *merciless_combat
*= 1 + 0.125*disease_count
crit% += subversion * 3% + rime * 5%
crit_bonus_multiplier *= 1.0 + 0.15 * GoG

That's what it looks like now


Okay, I did some testing to look into this. Here are my results:

The glyph appears to be additive with the base multiplier of the ability, the tooltip after applying the glyph reads: "Deals 100% of weapon damage plus 584, increased by 12.5% for each disease". Disease multipliers appear to be additive, so the formula for damage after disease application should be: 1.25*Weapon+730. Merciless combat appears to be additive with the disease multiplier, so it execute range should yield: 1.37*Weapon+800. Finally, tundra stalker appears to be multiplicative on top of all that, so with that applied the final formula for oblit damage should be: 1.58*Weapon+920.

Also looked into Frost Strike and it looks like many of it's multipliers are additive as well. My testing indicates that everything but tundra stalker is additive, so the frost strike formula should look something like: (0.6+165)*(1+0.1+0.2+0.12)*1.15 = 0.98*Weapon+269 (assuming execute range).

Looking at these formulas, it looks like like Oblit should be comparable damage to Frost Strike after armor reduction (and noticeable more if you have any arpen at all), which is what occurs on live but the sim puts it at about 66% of FS's damage.

ginger booty get on with yo bad self


#57 tripp6sic6

tripp6sic6

    Von Kaiser

  • Members
  • 31 posts

Posted 03 December 2009 - 08:24 PM

Death's choice determines the proc by the the AP the trinket has: 288 vs 256, both proc independantly

What I'm saying is that I don't believe that this is working. I'll give you an example of what I mean:

Profile 1:
#!./simc 

patch=3.3.0
threads=2
#calculate_scale_factors=1
#weapon_speed_scale_factors=1
iterations=20000
html=DeathKnight_T9_17_00_54.html
output=DeathKnight_T9_17_00_54.txt
death_knight=DeathKnight_T9_17_00_54
level=80
race=tauren
talents=http://www.wowhead.com/?talent#jfVMVZZfMbI0cdg0ux0uIqo
glyphs=the_ghoul/dark_death/icy_touch/horn_of_winter/pestilence/raise_dead
actions=flask,type=endless_rage
actions+=/food,type=dragonfin_filet
actions+=/presence,choose=blood
actions+=/snapshot_stats
actions+=/auto_attack
actions+=/raise_dead
actions+=/plague_strike,blood_plague<=2
actions+=/icy_touch,frost_fever<=2
actions+=/blood_strike,death=0
actions+=/scourge_strike
actions+=/summon_gargoyle,time<=20
actions+=/summon_gargoyle,bloodlust=1
actions+=/death_coil
actions+=/horn_of_winter
actions+=/empower_rune_weapon
ranged=sigil_of_virulence
# Gear Summary
gear_strength=1843
gear_agility=118
gear_stamina=1706
gear_intellect=20
gear_spirit=20
gear_attack_power=901
gear_expertise_rating=214
gear_armor_penetration_rating=94
gear_hit_rating=264
gear_crit_rating=1085
gear_haste_rating=342
gear_armor=14987
meta_gem=relentless_earthsiege
tier9_2pc_melee=1
tier9_4pc_melee=1
trinket1=deaths_verdict,stats=256ap
trinket2=deaths_verdict,stats=288ap
main_hand=dualblade_butcher,weapon=axe2h_3.60speed_811min_1217max,enchant=rune_of_the_fallen_crusader
Result:
9626 DPS and one listing of Death's Verdict under the Dynamic Buffs section.

Profile 2:
#!./simc 

patch=3.3.0
threads=2
#calculate_scale_factors=1
#weapon_speed_scale_factors=1
iterations=20000
html=DeathKnight_T9_17_00_54.html
output=DeathKnight_T9_17_00_54.txt
death_knight=DeathKnight_T9_17_00_54
level=80
race=tauren
talents=http://www.wowhead.com/?talent#jfVMVZZfMbI0cdg0ux0uIqo
glyphs=the_ghoul/dark_death/icy_touch/horn_of_winter/pestilence/raise_dead
actions=flask,type=endless_rage
actions+=/food,type=dragonfin_filet
actions+=/presence,choose=blood
actions+=/snapshot_stats
actions+=/auto_attack
actions+=/raise_dead
actions+=/plague_strike,blood_plague<=2
actions+=/icy_touch,frost_fever<=2
actions+=/blood_strike,death=0
actions+=/scourge_strike
actions+=/summon_gargoyle,time<=20
actions+=/summon_gargoyle,bloodlust=1
actions+=/death_coil
actions+=/horn_of_winter
actions+=/empower_rune_weapon
ranged=sigil_of_virulence
# Gear Summary
gear_strength=1843
gear_agility=118
gear_stamina=1706
gear_intellect=20
gear_spirit=20
gear_attack_power=901
gear_expertise_rating=214
gear_armor_penetration_rating=94
gear_hit_rating=264
gear_crit_rating=1085
gear_haste_rating=342
gear_armor=14987
meta_gem=relentless_earthsiege
tier9_2pc_melee=1
tier9_4pc_melee=1
trinket1=deaths_verdict,stats=256ap
trinket2=deaths_choice,stats=288ap
main_hand=dualblade_butcher,weapon=axe2h_3.60speed_811min_1217max,enchant=rune_of_the_fallen_crusader
Result:
10013 DPS and one listing for Death's Choice and one listing for Death's Verdict under the Dynamic Buffs section.

It doesn't appear that, when they are named the same, that two separate procs occur.

#58 Starfox

Starfox

    King Hippo

  • Members
  • 578 posts

Posted 03 December 2009 - 08:26 PM

Okay, I did some testing to look into this. Here are my results:

The glyph appears to be additive with the base multiplier of the ability, the tooltip after applying the glyph reads: "Deals 100% of weapon damage plus 584, increased by 12.5% for each disease". Disease multipliers appear to be additive, so the formula for damage after disease application should be: 1.25*Weapon+730. Merciless combat appears to be additive with the disease multiplier, so it execute range should yield: 1.37*Weapon+800. Finally, tundra stalker appears to be multiplicative on top of all that, so with that applied the final formula for oblit damage should be: 1.58*Weapon+920.

Also looked into Frost Strike and it looks like many of it's multipliers are additive as well. My testing indicates that everything but tundra stalker is additive, so the frost strike formula should look something like: (0.6+165)*(1+0.1+0.2+0.12)*1.15 = 0.98*Weapon+269 (assuming execute range).

Looking at these formulas, it looks like like Oblit should be comparable damage to Frost Strike after armor reduction (and noticeable more if you have any arpen at all), which is what occurs on live but the sim puts it at about 66% of FS's damage.

Oh, glyph being additive with the 80% should be a quite noticable difference, and when I did some excessive testing with 3.3 scourge strike I noticed that pretty much everything is additive:
bone shield, blood pres., desolation, black ice
Fixed quite abit with obliterate + glyph + sigil of awareness interaction in SC in r4052
Also: Is the bonus multiplier 12.5% per disease?
I mainly test on PTR and there it is only 10% per disease, although the tooltip clearly states 12.5%

@tripp6sic6:
Oh, thanks, we'll look into that
Hello.
Light the fuse.
For all my homies.
Do not run, we are your friends.
SimulationCraft Druid Guy

#59 dedmonwakeen

dedmonwakeen

    Bald Bull

  • Members
  • 1,302 posts

Posted 04 December 2009 - 02:38 PM

Also, I found a bug that I don't think is DK specific (I will test to make sure), but it seems that profiles that use the Heroic and non-Heroic versions of Death's Choice/Verdict will only use one proc. The workaround I have been using is to use the alternative name for one of the trinkets.


This is fixed in r4057 and will be included in the next release.

#60 Darkside

Darkside

    I find your lack of faith disturbing.

  • Allied Members
  • 9,820 posts

Posted 04 December 2009 - 05:24 PM

Oh, glyph being additive with the 80% should be a quite noticable difference, and when I did some excessive testing with 3.3 scourge strike I noticed that pretty much everything is additive:
bone shield, blood pres., desolation, black ice
Fixed quite abit with obliterate + glyph + sigil of awareness interaction in SC in r4052
Also: Is the bonus multiplier 12.5% per disease?
I mainly test on PTR and there it is only 10% per disease, although the tooltip clearly states 12.5%


From what I could tell, it was 12.5% per disease on live.

ginger booty get on with yo bad self





0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users